It was found that three cracked control rod drive mechanism tubes allowed boric acid to leak from the reactor’s cooling system on to the vessel head.The corrosion left only 6.35mm of stainless steel to protect the top of the reactor vessel, normally shielded by 165mm of carbon steel. The corrosion did not go right through the head due to a thin layer of non-corrosive stainless steel cladding. During ultrasonic inspection tests, a second, much smaller, cavity was discovered near the original one.

The NRC has issued a "confirmatory action letter" to owners FirstEnergy, "confirming the company’s commitments regarding actions it will take to evaluate and resolve damage to a small area of the reactor pressure vessel head at the Davis Besse nuclear power station." The letter confirms commitments FirstEnergy has made to the NRC, which include the determination of the root cause of the degradation and the agreement to meet with the NRC to obtain restart approval.

FirstEnergy said that the repairs would keep the 935MWe unit offline for 60 to 90 days beyond the originally planned restart in late March – provided that the NRC gives them permission to do so. FirstEnergy expects it will cost $5-10 million to repair the RPV head. In a letter to investors, it also said that it would cost $10-15 million a month to buy replacement power while the reactor is offline.

A replacement vessel head, costing some $20 million, was ordered last year after circular cracks around the vessel head penetration nozzle were found at Oconee. The new reactor head is due to arrive in February 2004. FirstEnergy is also looking into the possibility of salvaging a used vessel head from a decommissioned reactor, should one be required sooner.

Following the discovery, the NRC ordered operators of all 69 operating PWR units to submit reactor safety information. The NRC asked for information "on the extent of recent visual examinations of reactor pressure vessel heads, as well as plans for conducting such inspections during outages this spring." Utilities were given two weeks to respond.

Some preliminary answers were given in a report prepared by the materials reliability programme at EPRI. A more detailed survey is about to be submitted to the NRC. According to the report, about one third of the units were completely checked for corrosion during their most recent outage. For the remaining plants, most utilities qualified their answers with detailed descriptions of why they were still confident that they do not have a similar problem. Director of engineering at the Nuclear Energy Institute Alex Marion said raw data from the utilities suggests they would have found evidence of boric acid corrosion, had it existed at their units.